The fundamental premise of business intelligence has traditionally been “to provide the right information to the right people at the right time and at the right cost.” While this statement is irrefutable, it would be more accurate if we changed the word “information” to “actionable information.”
The fact is, most users have difficulty identifying critical metrics within the reports they receive. This is a major problem. It is important to clearly convey performance activity to decision-makers of organizations in order to help them make data-driven, actionable decisions.
Are users at fault for their difficulty identifying critical metrics? Not necessarily. If most students are failing a test, maybe something within the test is the cause of the problem rather than the knowledge of the students. Similarly, maybe the metrics are not clearly communicated.
So, the question is, how can the metrics be communicated more effectively? Words and numbers may have failed in relaying actionable messages, but this is where effective dashboards are useful. A dashboard is an interface that presents information in an easy-to-understand and easy-to-relate, often graphical way, providing users with a lot of meaningful information at a glance. Dashboards can be an incredibly valuable and empowering tool for understanding critical business data in the metrics.
The metrics used in the dashboard vary by industry and business function, as well as by the type of decision-maker and level of skill and tool used. They can be used for many functions, including planning, resource allocation, budgeting/forecasting, reporting, monitoring, and analyzing as well as strategy setting.
Dashboards are made up of two main components:
Key performance indicators: An enterprise performance scorecard (see below) and dashboard are based on meaningful and well-defined KPIs, which show whether an organization is meeting its stated objectives. They apply to the performance of the organization as a whole, so if there are multiple divisions, it should be determined whether each division will also need a scorecard and a dashboard.
KPIs can be financial, customer, internal process and employee-related. Many KPIs are interrelated - they don’t stand alone. For instance, common examples of KPIs include debt to equity ratio, asset turnover, and profit margin - and all the three are interrelated. The nonfinancial KPIs should have a cause-and-effect relationship with the financial indicators.
Scorecards: A scorecard is a set metrics (or KPIs) that presents current performance data for a business process or for a strategic goal toward a target value. These indicators direct the business on key tactical objectives and goals. This will ultimately shape the vision and strategy for the organization for that particular time period.
Challenges and Pitfalls
There are many potential challenges of implementing scorecards and KPIs. These include maintaining balance across multiple perspectives; determining and developing a few, solid important objectives and metrics; ensuring a mix of cause-and-effect and decomposition metrics; and combining the multiple dimensions of a matrix organization (such as business units, functions, processes).
Dashboard implementations are considered faulty if the following conditions exist:
- Data has to be entered manually or “refreshing” the data proves very time consuming.
- Calculating and/or aggregating data is tedious.
- The user has to navigate multiple tools in order to answer a question.
- The dashboard design is poor (not user-friendly).
- Multiple logins are required to get to the right dashboard.
- It needs an extensive amount of user training. The software tool has to be intuitive, resulting into almost zero necessary training.
- It lacks executive sponsorship and funding.
- The user must wait 10 seconds or longer to retrieve information.
Types of Dashboards
There are several different types of dashboards, including strategic, tactical and operational.
Strategic dashboards can reflect enterprise-wide strategic goals, as well as corresponding KPIs. Features on this type of dashboard include global, external, trends and growth measures, all of which are related to or based on the Balanced Scorecard Methodology (see Figure 1). The data on these dashboards is highly summarized and presented graphically, without too many details. However, details should be available when someone wants to drill down. This type of dashboard requires updates less frequently.
Tactical (also called analytical) dashboards measure the business’s progress according to related trends, in accordance with each strategic initiative. Progress is measured against a preset goal, such as a budget or a certain target. Drilldowns reveal details and break down data for analysis. For example, they help determine why certain targets were not met and where a potential problem might be.
Operational dashboards monitor specific business processes, such as order processing and shipping. They are mainly used at the departmental level, where operations take place. Updates are tracked daily or weekly using real-time charts and reports, and detailed data is presented with strong analytical functionality in order to perform a root-cause analysis.
No matter which type of dashboard you’re implementing, the ideal placement of the dashboard is the enterprise portal, because this is the central point from which organizations can distribute various BI applications. The portal enables everyone to access their dashboards, documents, presentations, online discussions and other applications from a single location. Additionally, within the portal users can discuss observations and concerns identified from their dashboards and can propose ideals and potential solutions.
To properly implement a dashboard, it is important to enlist proven project management practices. Like any major project, there are both people and technology issues, and the people issues need to come first.
- Choose a sponsor and determine what is expected by him or her in return for the funding. Find out whether the sponsor has sponsored any other projects and if he or she has been steadfast when the going gets rough (for instance with deadlines, budget and “scopecreep”), because these problems are likely to creep in. Based on all of this, select a suitable sponsor and ensure funding is approved.
- Identify key users and determine their expectations and what they need to know.
- Gather an implementation team that will cover functions such as project management, business analysis, integration expertise, data warehouse expertise and metrics development, including KPIs and scorecards.
- Determine which data sources will be used for each dashboard.
- Identify main KPIs used in the dashboard.
- Determine how often the dashboard needs to be updated.
- Identify needed hardware and software.
- Determine whether a screen designer needs to be assigned to aesthetically organize the dashboard for details such as selecting colors and determining placement of charts and diagrams. Keep this person on your team, because he or she can play a role in maintaining the storage used by the dashboard (which will be excessive with all media converging on one dashboard) and in handling the the time requirements of transmitting the dashboard around the globe.
Dos and Don’ts of Effective Dashboards
Just because you have a dashboard doesn’t mean it’s an effective one. Here are some tips to create a dashboard that conveys actionable information to users.
- Know your audience and their interests: Determine which questions your dashboard will answer and for whom these answers are pertinent. Also, consider the variety of devices (laptop, mobile phones and mobile “gadgets”) through which to deliver the dashboard and what will be most accessible as well as useful to the greatest number of people.
- Develop accurate and consistent dashboards: Minimize the redundancy of the data that you are using for the dashboard. Most inaccuracy and inconsistency is due to redundantly stored data.
- Have up-to-date data: Not all data is produced or updated in real time. It is important to know what data is current and still relevant to the needs of your audience.
- Make the dashboard easy to read: Don’t clutter the screen. Instead, draw attention to the core content in the dashboard. Provide space for the audience to send questions or comments.
- Identify critical metrics (with critical KPIs): Determining which metrics are most pertinent to the users is essential. Also, business users should be capable of drilling down into the KPIs for lower level details.
- Think “dynamic”: Make the dashboard as interactive as possible, and make it easy to create content on an ad hoc basis. Identify approximately 25 percent of the metric categories that have remained on the dashboard for an extended amount of time that have provided some kind of constant support to the viewers. Consider moving (or removing) the categories that are used less frequently.
- Provide a customizable dashboard interface: Business users should be able to easily customize their dashboards without any assistance from IT. Also, provide a facility so that business users are able to create their own dashboards.
- Provide a printable dashboard: Most senior managers still want to see a hard copy.
- Use dashboard analytics with visualization:
- Import your most useful, frequently used and revealing data about your business today, this week, this month.
- Build a BI dashboard or a BI report with appropriate KPIs for people making the most up-to-date decisions.
- Publish the BI dashboard or a BI report on cloud servers. This will make the dashboard instantly accessible to a vast number of people at any time.
- Don’t extend the dashboard beyond one page or screen: The dashboard should present the most important results on a single page. When unnecessary information is not crammed in, the material is more understandable, and, as a result, the information is more actionable.
- Don’t present data that is dependent on other data: All data in the dashboard should be fully comprehensive; additional explanations should not be needed to understand it. Inventory data, for example, should be self-explanatory, and it should not be necessary to look up orders to relate to the inventory.
- Don’t provide only one level of data: In line with the drill-down notion, it’s important to present multiple layers of data. Your dashboard design should be able to drill down at least another two to three levels. Even if you don’t make that information presentable in the first version of the dashboard, the design and architecture should allow it to be easily expandable and accessible later.
- Don’t present metrics in a vacuum: There should be context in which to provide valuable insight into the presented data (for instance, high number of house foreclosures in the context of the collapsed real estate market). Use KPIs, goals, benchmarks or even some time series results to give relatable context. For example, if we are told that the revenue this quarter is $XXX million, it is a naked number with no context in which to analyze it. But if we know the revenue for the same quarter from the last two years, then we are able to compare the numbers and make important deductions about the data.
- Don’t expect your first dashboard to be your finest: Build and learn; there is a learning curve involved. Get into a “release” mentality – and announce it to the audience: develop a mock dashboard and consider it to be on training wheels, observe how it is used and received by your audience, and then build a real one. Notice the benefits of your dashboard, as those are assets that will help boost funding for your next dashboard project. Additionally, use software tools that are flexible and will enable you to make changes easily as you build your first dashboard.
- Don’t try to answer every question with one dashboard: A dashboard should not be a catchall. You may consider building a number of dashboards with various focal points instead of having only one with all important criteria jumbled into one dashboard. Those various dashboards could be accessible via a company portal on the intranet.
- Don’t have too many metrics: You are going to miss everything if you try to cover too many important performance metrics in one dashboard. Too many metrics will defeat the purpose of the dashboard because you won’t see the trees through the forest. Provide only a few critical metrics.
In today’s business world, data and data-driven decision-making are more important than ever. With the information provided in this column, you should have the necessary tips to implement an effective dashboard that will launch your organization into making better decisions to reach its strategic goals. I can confidently say that a dashboard is the new face of BI.
Editor's note: This is the seventh in a series of articles by Shaku Atre. Click on the titles to read the other recent articles: "Who in the World Wants to Stay Locked Up?"; "Who in the World Doesn't Want to Reach for the Clouds?"; "Who in the World Wouldn’t Want a Collaborative BI Architecture?"; "Who in the World Wants More Data?"; "Who in the World Needs a Data Warehouse?"; "Who in the World Wouldn’t Want to Evaluate BI Products?"; "Who in the World Needs a Hard Drive?"; and "Who in the World Wants to Just Be Structured?"
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access